Self-Composing Policies for Scalable Continual Reinforcement Learning.

Mikel Malagn, Josu Ceberio, Jos Antonio Lozano

Browse the full ICML paper archive.